Discussion prompt about AI systems failing discovery-oriented work by prematurely converging on answers.

Developer discovers AI text detectors are unreliable after personal testing.

A personal blog-style walkthrough of how LLMs work, criticized for poor structure and unclear audience.

Simon Willison covers OpenAI's Lockdown Mode feature.

LangChain repository listed as 'agent engineering platform' with no technical detail.

Large language models (LLMs) perform increasingly well on biology benchmarks, but it remains unclear whether they uplift novice users -- i.e., enable humans to perform better than with internet-only r...

Scaling multimodal alignment between video and audio is challenging, particularly due to limited data and the mismatch between text descriptions and frame-level video information. In this work, we tac...

Internet measurement faces twin challenges: complex analyses require expert-level orchestration of tools, yet even syntactically correct implementations can have methodological flaws and can be diffic...

Test-time training (TTT) with KV binding as sequence modeling layer is commonly interpreted as a form of online meta-learning that memorizes a key-value mapping at test time. However, our analysis rev...

Rapid progress in video models has largely focused on visual quality, leaving their reasoning capabilities underexplored. Video reasoning grounds intelligence in spatiotemporally consistent visual env...

In jurisdictions like India, where courts face an extensive backlog of cases, artificial intelligence offers transformative potential for legal judgment prediction. A critical subset of this backlog c...

In many real-world settings, such as environmental monitoring, disaster response, or public health, with costly and difficult data collection and dynamic environments, strategically sampling from unob...

In multi-agent IR pipelines for tasks such as search and ranking, LLM-based agents exchange intermediate reasoning in terms of Chain-of-Thought (CoT) with each other. Current CoT evaluation narrowly f...

Large language models (LLMs), and conversational agents based on them, are exposed to personal data (PD) during pre-training and during user interactions. Prior work shows that PD can resurface, yet u...

The average reward is a fundamental performance metric in reinforcement learning (RL) focusing on the long-run performance of an agent. Differential temporal difference (TD) learning algorithms are a ...

Operational rigor determines whether human-agent collaboration succeeds or fails. Scientific data pipelines need the equivalent of DevOps -- SciOps -- yet common approaches fragment provenance across ...

Generalized additive models (GAMs) offer interpretability through independent univariate feature effects but underfit when interactions are present in data. GA$^2$Ms add selected pairwise interactions...

We present Lifelong Scalable Multi-Agent Realistic Testbed (LSMART), an open-source simulator to evaluate any Multi-Agent Path Finding (MAPF) algorithm in a Fleet Management System (FMS) with Automate...

The long-standing vision of general-purpose robots hinges on their ability to understand and act upon natural language instructions. Vision-Language-Action (VLA) models have made remarkable progress t...

Despite rapid progress on coding agents, progress on their multimodal counterparts has lagged behind. A key challenge is the scarcity of evaluation testbeds that combine the complexity of software dev...

We present ROCKET, a training-free model compression method that achieves state-of-the-art performance in comparison with factorization, structured-sparsification and dynamic compression baselines. Op...

How can children acquire native-level syntax from limited input? According to the Poverty of the Stimulus Hypothesis (PoSH), the linguistic input children receive is insufficient to explain certain ge...

Large Language Models (LLMs) often provide chain-of-thought (CoT) reasoning traces that appear plausible, but may hide internal biases. We call these *unverbalized biases*. Monitoring models via their...

Large Language Models (LLMs) often rely on test-time scaling via parallel decoding (for example, 512 samples) to boost reasoning accuracy, but this incurs substantial compute. We introduce CoRefine, a...

OpenAI's first open-weight LLMs since GPT-2 (2019). Apache 2.0 license. Trained with RL and distillation from o3 and frontier internal models. GPT-oss-120B runs on single 80GB GPU; 20B runs on 16GB edge devices.

Cosmos Reason 2 is an open reasoning VLM enabling machines to see, understand, and act in the physical world. GR00T N1.6 is a vision-language-action (VLA) model for humanoid robots integrating egocentric camera streams, robot states, and language instructions into a unified policy.

Flagship reasoning model with adaptive tool-use -- intelligently invokes retrieval and code interpreter on demand during inference. Advanced test-time scaling via RL.

First open-source model supporting three video generation modes in one architecture: multi-subject reference image-to-video, audio-driven avatar generation, and video-to-video editing. Intelligent shot-switching for minute-level durations.

Flow and diffusion models produce high-quality samples, but adapting them to user preferences or constraints post-training remains costly and brittle, a challenge commonly called reward alignment. We ...

Adapting large pretrained models to new tasks efficiently and continually is crucial for real-world deployment but remains challenging due to catastrophic forgetting and the high cost of retraining. W...